When the Interface Starts Thinking: Designing for Agency in the Age of AI
AI no longer just executes instructions. It interprets context, anticipates intent, and quietly reshapes user journeys before a single click occurs. When systems begin to act in this way, familiar UX assumptions start to wobble, and design responsibility shifts in subtle but important directions.
This talk explores what it means to design responsibly when AI systems take on quasi-autonomous roles at the interface. Drawing on work in deceptive design, platform governance, and emerging AI regulation, it offers a practical framework for distinguishing between AI that supports users, AI that steers them, and AI that crosses the line into manipulation or unfairness.
Bringing together law, behaviour, and product design, the session explains why contemporary AI creates new categories of risk for designers, translates key elements of Europe’s evolving digital rulebook into everyday design guidance, and presents design approaches that preserve user agency and trust while still embracing AI’s